Hypoxia-induced circSTT3A enhances serine synthesis and promotes H3K4me3 modification to facilitate breast cancer stem cell formation

Hypoxia is a key feature of tumor microenvironment that contributes to the development of breast cancer stem cells (BCSCs) with strong self-renewal properties. However, the specific mechanism underlying hypoxia in BCSC induction is not completely understood. Herein, we provide evidence that a novel hypoxia-specific circSTT3A is significantly upregulated in clinical breast cancer (BC) tissues, and is closely related to the clinical stage and poor prognosis of patients with BC. The study revealed that hypoxia-inducible factor 1 alpha (HIF1α)-regulated circSTT3A has a remarkable effect on mammosphere formation in breast cancer cells. Mechanistically, circSTT3A directly interacts with nucleotide-binding domain of heat shock protein 70 (HSP70), thereby facilitating the recruitment of phosphoglycerate kinase 1 (PGK1) via its substrate-binding domain, which reduces the ubiquitination and increases the stability of PGK1. The enhanced levels of PGK1 catalyze 1,3-diphosphoglycerate (1,3-BPG) into 3-phosphoglycerate (3-PG) leading to 3-PG accumulation and increased serine synthesis, S-adenosylmethionine (SAM) accumulation, and trimethylation of histone H3 lysine 4 (H3K4me3). The activation of the H3K4me3 contributes to BCSCs by increasing the transcriptional level of stemness-related factors. Especially, our work reveals that either loss of circSTT3A or PGK1 substantially suppresses tumor initiation and tumor growth, which dramatically increases the sensitivity of tumors to doxorubicin (DOX) in mice. Injection of PGK1-silenced spheroids with 3-PG can significantly reverse tumor initiation and growth in mice, thereby increasing tumor resistance to DOX. In conclusion, our study sheds light on the functional role of hypoxia in the maintenance of BCSCs via circSTT3A/HSP70/PGK1-mediated serine synthesis, which provides new insights into metabolic reprogramming, tumor initiation and growth. Our findings suggest that targeting circSTT3A alone or in combination with chemotherapy has potential clinical value for BC management.